Project Task #9041
Project WP #665: WP7 - Supporting Blue Environment: VREs Development [Months: 7-30]
|Status:||Closed||Start date:||Jun 26, 2017|
|Assignee:||Emmanuel Blondel||% Done:|
|Lead beneficiary:||12 - CLS||Participants:||5 - FAO|
R process will:
• rely on WFS for grabing validated feature (an aquaculture farm or set of cages for the farm, to be clarified with CLS @firstname.lastname@example.org )
• upload the feature as shapefile in GeoServer public & publish layer based on it
• generate & publish ISO 19115/19139 metadata to Geonetwork
To be plugged as SAI/Dataminer algorithm
#3 Updated by Emmanuel Blondel over 2 years ago
@email@example.com which prod Geonetwork can I use here? There is no instance for the aquaculture atlas (I didn't request it because Lino pointed out limited hardware resources). Can I use (at least to start) this geonetwork: http://geonetwork.d4science.org/geonetwork ? If yes, what are the credentials?
#4 Updated by Fabio Sinibaldi over 2 years ago
Hi @firstname.lastname@example.org, to know what resources are assigned to a certain scope, you can always use our gui here : https://services.d4science.org/infrastructure-monitor
I'm also producing a wiki documentation on SDI-Service available at https://gcube.wiki.gcube-system.org/gcube/Interaction_with_SDI-Service . It's not complete (I'm gonna publish it on the portal as soon as it's ready), but the "SDI" interface described there should be enough for you to gather GeoNetwork credentials.
#6 Updated by Emmanuel Blondel over 2 years ago
- % Done changed from 0 to 60
- Participants 5 - FAO added
The data publication part of the R script is ready and functional. It includes the following workflow:
* grab Farm data from WFS
* push it as shapefile in public workspace
* create/publish the layer if not existing, otherwise update featureType (for bbox update, in case)
* create/publish the cluster layer if not existing, create it by pushing the individual farm shapefile (as separate one), otherwise, grab all features other than target individual farm: if empty collection, create it from individual farm, otherwise append new farm. The initial approach envisaged for having the cluster layer was to rely on GeoServer aggregates or "layergroups", but the latter doesn't support WFS (since grouped layers have potentially heterogenous schemas), so i've switched to a shapefile merging & update process.
* to append metadata creation & publication/update for both individual farm layer & clustered layer
* to test data validation with incoming data from Greece (with Greek characters), to see if similar hack as for #9059 should be added (to make Geoserver treating well encoding)
* to deploy the script as DM service
#10 Updated by Emmanuel Blondel over 2 years ago
@email@example.com and @firstname.lastname@example.org See here http://geonetwork.d4science.org/geonetwork/srv/en/main.home some metadata how they are created/published with the data validation process. In the search type "AAPS". ASAP the script will be deployed as service.
There is room to improve the metadata from content viewpoint but also legal aspects (Anton, yet other material for people working on legal metadata elements). For content, i discussed with CNR in order to be able to fetch user profile information, see #9253 for more information. I've put in place a Lineage information where process steps are described (i copy @email@example.com as well as he might be interested to get inspiration from this lineage stuff), here as well we may inherit information. e.g. for the data production, we could link to another metadata to be associated with staging products.
#14 Updated by Emmanuel Blondel over 2 years ago
- % Done changed from 0 to 90
Process deployed here: https://i-marine.d4science.org/group/rprototypinglab/data-miner?OperatorId=org.gcube.dataanalysis.wps.statisticalmanager.synchserver.mappedclasses.transducerers.AAPS_PUBLIC_PUBLISHER
The process is not fully operational, working the first time, then fails due to issue highlighted in #9251 (impossible to perform metadata GET/UPDATE). Once #9251 is solved, i will perform final testing in link to WFS-T edition with Greek farm names to tackle possible encoding issues.